On the Efficacy of Sampling Adapters

Clara Meister, Tiago Pimentel, Luca Malagutti, Ethan Wilcox, Ryan Cotterell


Abstract
Sampling-based decoding strategies are widely employed for generating text from probabilistic models, yet standard ancestral sampling often results in text that is degenerate or incoherent. To alleviate this issue, various modifications to a model’s sampling distribution, such as top-p or top-k sampling, have been introduced and are now ubiquitously used in language generation systems. We propose a unified framework for understanding these techniques, which we term sampling adapters. Sampling adapters often lead to qualitatively better text, which raises the question: From a formal perspective, how are they changing the token-level distributions of language generation models? And why do these local changes lead to higher-quality text? We argue that the shift they enforce can be viewed as a trade-off between precision and recall: while the model loses its ability to produce certain strings, its precision rate on desirable text increases. While this trade-off is not reflected in standard metrics of distribution quality (such as perplexity), we find that several precision-emphasizing measures indeed indicate that sampling adapters can lead to probability distributions more aligned with the true distribution. Further, these measures correlate with higher sequence-level quality scores, specifically, Mauve.
Anthology ID:
2023.acl-long.80
Original:
2023.acl-long.80v1
Version 2:
2023.acl-long.80v2
Volume:
Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
Month:
July
Year:
2023
Address:
Toronto, Canada
Editors:
Anna Rogers, Jordan Boyd-Graber, Naoaki Okazaki
Venue:
ACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
1437–1455
Language:
URL:
https://aclanthology.org/2023.acl-long.80
DOI:
10.18653/v1/2023.acl-long.80
Bibkey:
Cite (ACL):
Clara Meister, Tiago Pimentel, Luca Malagutti, Ethan Wilcox, and Ryan Cotterell. 2023. On the Efficacy of Sampling Adapters. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 1437–1455, Toronto, Canada. Association for Computational Linguistics.
Cite (Informal):
On the Efficacy of Sampling Adapters (Meister et al., ACL 2023)
Copy Citation:
PDF:
https://aclanthology.org/2023.acl-long.80.pdf